Preview by Point Differential: 2014 NBA Finals

In the media, LeBron’s rival was supposed to be Kobe Bryant. No one expected LeBron to face Tim Duncan and the Spurs for a third time in the Finals, but maybe we should be appreciative.

Duncan is regularly labeled as the best power forward ever, but LeBron is on the path to clinching the title of best small forward ever. How often have two players who are the best at their respective positions met in the Finals? Magic Johnson and Michael Jordan have a lock on their own positions, and they met in the 1991 Finals in a passing-of-the-torch moment. If Kareem is the representative for centers, that’s the only other time two players who are considered the best have faced each other in the Finals, excluding the times Duncan and LeBron have met. (Kareem and Magic shared a team, of course; it’s that kind of star-power that ignited the famous 1980’s Celtics-Lakers rivalry.)

The Finals are here. Duncan and LeBron are already legends. What happens next?

Measuring team strength in a player’s league

One of the laziest assumptions in modern sports analytics is that a team’s strength is just that: look at the team’s point differential, offensive or defensive ratings, their win %, etc. and one has a good idea at how strong a specific team is. But teams aren’t one unit; they’re not one discrete element. It’s a collection of players, and in the NBA even one single player can drastically change the results.

Nevertheless, team strength is a good starting point. A team’s rating was found via a regression equation with point differential and every game up to the finals. It also accounts for homecourt advantage and the effect of back-to-back games. (This system is similar to basketball-reference’s SRS, but it includes more information.) Based on this form of adjusted point differential, and remember this includes the playoffs, the Spurs appear to be much better than the Heat.

San Antonio Spurs:
+8.70 rating (63.1 wins with neutral schedule)
Miami Heat:
+5.07 rating (54.8 wins with neutral schedule)

Throw in homecourt advantage and the Spurs are a little over 3 points better per game. However, that includes games where Wade was injured or resting. How strong are the full strength Heat (with all three All-Stars)? And to make this fair, how strong are the Spurs when Duncan, Parker, and Ginobili play? (In the regression, the Spurs and Heat are split into two teams each: one where every “big 3” player is playing and one with at least one missing.)

San Antonio Spurs (with Duncan, Parker, and Ginobili playing):
+9.71 rating (65.1 wins with neutral schedule)
Miami Heat (with James, Wade, and Bosh playing):
+5.52 rating (56.0 wins with neutral schedule)

San Antonio’s advantage increases when all of their stars are playing. They have a roughly 80% chance at winning the first game, and even in Miami they have a 50-50 shot at winning.

A Miami fan would counter that the Heat were “coasting” through the regular season and with LeBron James in the mix, they will always have the better chance at a title. However, the Heat are mere amateurs at coasting, as their starters played hundreds of minutes less than San Antonio’s starters, yet the Spurs still led the league in both wins and point differential. The series last season shouldn’t be used as a counter-example because the Heat had more wins and a higher point differential; and they won the series by a hair (Kevin Pelton found that the 2013 Spurs were as close to the title as any team has ever been with odds peaking at 99.3% in game 6.)

Why point differential matters

Coming off two titles in a row isn’t an excuse for playing worse. The two worst championship Bulls teams, arguably, were each at an endpoint of a three-peat, and the 2002 Lakers were the worst iteration of the Shaq-Kobe Lakers as well. However, in the Finals (or the conference finals for the Lakers, since the east was so weak that season) those teams outscored their opponents by 2.4 points per game more than you’d expect, given both team’s strength (basketball-reference’s SRS) and homecourt advantage. They completed the three-peat titles by playing better than was indicated by the regular season. Shaq and Jordan raised their games against the opposition.

But history is littered with failed three-peat adventures. Since the Bill Russell-era, there has been seven opportunities, and there are only three successes with four failures (Kobe-Gasol in 2011, Hakeem in 1996, Thomas-Dumars-Rodman in 1991, and Magic-Kareem in 1989.) The back-to-back champions who tried for a third title and failed have crashed hard, losing on average by 8.6 points per game in their final series.

The problem? Basically, they all got old, and that’s an issue with the Heat too: their role players, like Battier, Haslem, and Allen, are ancient in NBA terms. The Spurs are known as the grey-hairs, but at least they have a cache of young talent and a strong bench.

The Spurs look like likely champions based on point differential, but does it even matter? The yearly leaders for basketball-reference’s SRS, adjusted point differential, have won 35 titles out of 67 possible (52%), which isn’t a bad track record given how the field is open to multiple teams. This isn’t even corrected for teams with injuries, like OKC in 2013 (Miami had the second best point differential that season.)

Using SRS for every Finals since the merger in 1977, 25 out of 37 (67.6%) Finals were won by the team with the higher SRS. Since 1996, that figure improves to 15 out of 18 (83.3%.) Miami has been a part of the only three series since ’96 where the winner had the lower SRS. We can’t ignore it. Regular season point differential is important even at this stage in the playoffs, and I included playoff games for my own ratings. San Antonio is simply better.

The future is not the past

As useful as a team’s point differential is, it can break down because a team is not a single element. Beating Miami isn’t impressive if LeBron isn’t there. However, rating each team based on how valuable their individual players are doesn’t help Miami either.

The highest performing publicly available individual rating system, real plus-minus, still gives a major edge to San Antonio. One also can’t argue that Miami will win because it’s a better match-up for them. San Antonio has gotten better playing “smallball” and experimented with Bonner starting. If Miami starts a non-entity on offense like Battier, Bonner has somewhere to hide. Leonard also appears to be the best defender for LeBron James. The two teams split the season series with San Antonio, but the Spurs blew them out by 24 points one of those games (note: each “big three” was intact for both games.)

People assume Miami will win in a close series because of what happened last year, but it’s not 2013 anymore. Interestingly, it wasn’t too long ago that LeBron was the MVP who couldn’t win the title, and now he’s the MVP we assume will win the title — perceptions change quickly.

But the NBA is still unpredictable. The future will not be like the past. We estimate team strength based on recent results, but it’s not with 100% certainty. Wade was terrible in the 2013 playoffs, and he’s played like an All-Star this year. Tony Parker is battling injuries and will play through pain.

Even though the Spurs are clearly a better team, they only have a roughly two-thirds chance at winning the title (varying based on which team ratings you use.) That’s the difference between stating that a team has the highest probability of winning and stating that a team will win. The analytics world knows the stochastic nature of sports. No one can say who will win, despite the importance of the occasion.

It’s LeBron v. Duncan, round 3. Both are legends, both are among the top ten players ever. Now what?

Justin’s Prediction: Spurs in 5

Quantcast